Slashdot Mirror
Transmeta Needs Microsoft
An anonymous reader writes "Faced with dwindling sales, it looks like Transmeta
needs Microsoft's new tablet PC to survive." Or, if not Microsoft, some company who can spark the long-overdue tablet-computing revolution.
The best tablets, came from Sumeria.
If you get an error, type "OVERRIDE" or "SECURITY OVERRIDE" and then try the optimize command again.
And if Transmeta makes the chips, then that is double cool.
I personally don't care WHO makes the damn things, as long as someone makes them, and gets them out there for reasonable prices!
Supply and demand....but where's the demand?
They're so far behing on mfg technology that they're screwed. AMD or Intel can take a .13 micron part, underclock it, under-volt it and spank em silly (while getting many more chips per wafer).
Holy crap! If you hadn't pointed that out we never would have realized it.
and I pronounce "Microsoft" "Microsoft".
Slashdot: Failed Car Analogies. Amateur Lawyering. Anecdote Battles.
Fueling the frenzy was one of its lead software engineers, Linus Torvalds, who was already famous for developing the core of the Linux operating system
Linus was one of the lead software engineers for this company, and yet it needs Microsoft to keep it from flopping? Perhaps MS can do the hardware and Linus the software, which might actually make a good product.
What we need is linux open-sourceness with Microsoft marketing
Transmeta to create "profit morphing" technology to turn an unprofitable business plan into to a profitable one.
They will roll out newly patented "micro profit opcodes" to soak up Microsoft money at the molecular level.
If your business model needs MS, then you're already done.
Yeah, because NO ONE has EVER made a lot of money hitching themselves to the Windows platform... (*cough*Quicken, Visio, Norton, McAfee, Innumerable Games, etc, etc, etc*cough)
Sheesh, dude, get a clue. Microsoft is dominant exactly BECAUSE they make it easy for people to develop for the platform and make lots of money. Review the history of OS/2 to see what happens when you rape the developers.
Sometimes it's best to just let stupid people be stupid.
Transmeta should have realized a long time ago that they couldn't break into the laptop market (which is where they seem to have been trying to go). What Intel doesn't monopolize, AMD jumps on with greater resources than Transmeta could hope for (Hammer notwithstanding).
The MS Tablet PC could be the best thing to happen to Transmeta. MS isn't exactly happy with the major box brands offering *nix, and inside rumours say debate over whether security should be hardware or software have put a good dent in MS/Intel relations. By going Transmeta, MS can get a good price on a suitable processor (not megapowered like the portable P4's, but perfect for the job), which means more tablets running MS software, they get a bigger say in the design of the tablet, and the poke Intel in the ass and say "Don't push your luck, big guy, we don't need you as much as you want to believe."
And in the end, the result is just as good for open source on the tablet platform, because cheap tablets with a big company behind them will get a strong push into the marketplace, give OpenSource developers a reason to write for the tablet.
I say, Go Transmeta!
(hey, wasn't that a catchphrase from an 80's kids show?)
If I knew the wedgies I gave you back in 6th grade would have resulted in this . . . I might have taken a moments pause.
I wonder if that has much to do with the US way of life - we see almost exact same trend in automobile and home appliance industries. People in the States like to drive muscle cars, SUVs, full sized cars, who guzzle gas like crazy. In Europe, the trend is reverse, smaller, more economic cars running on electricity or natural gas and well esablished. Here in the US, they all seem to be either developing very slow or even failing.
Same thing with the dryers and washers. Europe in Japan goes for making them more energy efficient and smaller (due to space constraints, mainly), where here in the US we dont see much of a move away from the full sized washers.
Because of that, I would think the quoted statement could very true.
"If your business model needs MS, then you're already done."
Tell that to Gateway and Dell.
....maybe it's not in the cards?
The posts in the thread already mention that the Carusoe is a niche chip- from what I've seen, it's gone horrifically under-utilized: a chip that could hypothetically be a power pc, mips, sparc, etceteras is nothing other than a super-low power x86?
A tablet PC might be fine for some people- If the input is pressure sensitive, it would be great for the graphics field- but these really don't seem to be much more than big PDAs or totally integrated, one-piece laptops.
What, exactly, is a tablet good for that a PDA or laptop *isn't* ? I need quick access to photoshop and apache practically everywhere I go (freelance web designer with a powerbook)- a PDA is useless for me, and the tablets I've seen don't run my OS of choice or seem to do anything I might need.
Someone clue me- what market are tablets actually *aimed* at? A laptop is perfect for my needs, and a PDA works great for many people I know.
If people don't need a thing, or can't find a use for it, then the only people that are going to buy the device are gadget hounds- which, in all honesty, don't seem like enough of a market segment to keep a niche industry like this afloat.
Intel can afford to sell chips less expensively than it normally would in order to gain a foothold in a given market--and it has proven its willingness to engage in price wars.
...
This is the crux of the article, predatory pricing: airline seats, xboxs, OSs, etc. Sell the low-margin product at a loss to sell 5 high-grossing products for an AVERAGE price greater than your competitor. Even if the tablet PCs are a hit, they'll get squeezed when Intel wants that market share. So one foot in the grave at this point
"This isn't a study in computer science, its a study in human behavior"
The only resin Quicken exists is b/c MS lets them exist.
Oh man, look at this history of Quicken. Microsoft tried damn hard to kill Quicken. Damn hard. In fact, they were going to just give and buy out Quicken, but the DOJ blocked the buyout.
Don't you think they would just bundle Money with the OS if they wanted to stomp out Quicken?
Sure, if they wanted to screw themselves in the process. The point isn't to destroy Quicken, the point is to maximize profit. And giving away "Money for nothing" (heh) doesn't maximize profit. It's hard to start selling something again after you give it away.
If you think MS plays nice-nice for the benefit of the other company, then think again.
Microsoft is smart enough to realize that the biggest advantage they have is their range of applications. That's why they treat their developers so well.
On top of that, I think MS now owns Viso.
That's called being a successful Windows developer. If only I could create product that Microsoft would want to pay 100s of millions of dollars for. :)
Sometimes it's best to just let stupid people be stupid.
I was really interested in getting a Sony Picturebook. This was about 4-5 months ago, when the latest ones had not yet hit the United States. I asked a client of mine, who is Japanese, to get me pricing. He obliged, but only after warning me about the Transmeta processor. "It doesn't work well when you try to run multiple applications," he said. "Everyone says it's slow."
I asked him who had told him that. He said it was the Sony rep at the store where he bought his Vaio. Uh-oh.
I knew a Transmeta 867MHz processor wouldn't perform as well as an Intel 867MHz processor, but I did some digging and was shocked to figure out how much slower it really is. Check out these benchmarks from Tom's Hardware. The Transmeta 600MHz processor got stomped by a "vintage" PII/366MHz notebook. That's terrible.
To me, small size and battery life rank higher on my list than pure performance. Still, the Transmeta processors run so slowly that the only way I could justify buying one is if they had 5+ hours of battery life. But they don't -- the PictureBook is only advertising 2.5 hours of battery life. Compare this to the (admittedly larger) 3.7-pound IBM X30, where Walter Mossberg put one through the grinder and got 3 hours and 29 minutes of battery life. IBM is claiming 5+ hours in BatteryMark for the same laptop.
Transmeta did one thing, and that was to get Intel turned on to the fact that consumers want good battery life in notebooks. I think the quote from the article puts it best: "Intel's focus on battery life happened because Transmeta pressured them into it... forced them to do something different. The good news is you've got a giant to acknowledge you but the bad news is you've woken the giant."
Right now, the giant is still stomping Transmeta, and I doubt that tablet PCs will really put Transmeta back in the running. Whatever Transmeta can come up with, Intel has proven that they can match. Transmeta might make initial inroads, just like they did on subnotebooks, but eventually Intel will again wake up, and this time I don't think Transmeta will survive.
Simpli - Your source for San Jose dedicated servers and colocation!
Put the gun down! I was only kidding!
The number of humans that can outcode GCC is vanishingly small, and even smaller when you have to do keep track of all sorts of parallel dependencies and such. If I had to guess, I would say their problem is that they don't have the necessary capital to do the software development they would need to make this idea really fly. This seems like more of a research project than a business plan, maybe if they enlisted more support from the community by openning things up GPL style, it might have a chance.
If you are going to make chips, you should concentrate your investment on designing and making the chips, not the software. I'm sure they could do another design cycle in more modern fab technology and get much more speed (or lower power if that is the priority).
Also, my impression is that they aren't sharing a lot of board level IP either. There are lots of applications that are getting StrongArm and other low power processors that would be an ideal market for them, but these are all small companies without a lot of bucks to risk with new designs. OTOH, if you give them a basic design that uses your chips, and sell cheap prototyping parts and support gear, your going to get a lot more inquiries. Would this be enough to turn it around for them? I don't know, but would you bet your company on Billy G. deciding to endorse tablets? Even if they are successful, the danger is that Intel can swoop in and steal the market. You're much better off staying small and concentrating on the emerging niches. Specifically, the embedded market where Linux is already a good fit and doing well.
Too bad, though, it was a good idea, but the timing was bad. They still have some investment money that hasn't been spent, but it would be difficult to change direction now.
My wife did a course on Unix at UC-Berkeley. AOL, PC, MS Office were required for that course. This is the success story of AOL-Intel-MS which managed to sell themselves to a Unix student at UCB. Transmeta is just not big enough to sell themselves in a crowded market all by itself. Linux will not help them. They will need Microsoft support (who else can help them in x86 market). Most likely, even MS will not help them. They will exploit Transmeta to force Intel and AMD to come up with low-cost technology for tablet-pc that they won't to push through consumer's throat.
Except for a few limited fields, tablet PC's are really not ideal. They are well suited to any field where you need some portability, and expect to be doing more data viewing than entry. The field that is most obviously well suited to this is the medical field. There you have X-rays, tests, etc, that can all be loaded from a central server and you only need to make small annotations. You need some portability, but not much.
The problem is that not many fields really fit that bill. You usually want either total portability or very limited portability (thus making the laptop or PDA a good choice). Furthermore you sacrifice a lot of your ability to do data entry on a tablet PC with only a limited gain in portability. A well designed PDA is actually far superior for data entry because I can use a thumb keyboard and enter data quickly. A tablet requires one hand to hold the thing, and then the other to do data entry.
I've used them and I've found that, for the most part, they are solving a problem that I've never had. I suspect that it's a problem very few people have, so except in a few niche fields tablets aren't going to be a big thing. So buhbye transmeta.
This sig has been temporarily disconnected or is no longer in service
I've said it before and I'll say it again. Tablets will only be successful in vertical markets. Sure, they may sell some at first because of the GeeWhiz factor, but writing on the screen is simply less efficient than typing. If Transmeta and/or Microsoft is/are pinning their futures on tablets, then a rough future it/they will have.
What they needed then and need now is a comprehensive, grass-roots developer support program and to get humble quick about the power and performance.
It's a nifty freaking technology, and it was a LOT of fun to play with. Did you realize that the Crusoe during the launch party was the fastest native picojava bytecode processor in the world? Did you realize that had they exposed an interface to CMS (code morphing software), that someone out there could have written a PowerPC personality for it, allowing it to run both x86 and PPC apps at the same time? Can you imagine what else you could do with access to this incredibly powerful, real-time, back-end compiler? Did you realize that you could decode, issue and retire two integer ops and an MMX or fp op on every cycle, the same decode rate as a modern Athlon and a faster decode rate than the P4? Did you realize that all the tech is in place to allow you to download CMS upgrades that vastly improve performance? Did you know they have a perf monitoring tool that puts vtune to shame? Did you know that using gdb connected as a cross-debugger, you can hit "ctrl-C" in an NMI handler or anywhere else and get a complete dump of the internal processor state, including numerous perf statistics?
Tip of the iceburg for the current core, and their next generation architecture (TM8x00) is SO MUCH COOLER. Like hella-cool with chocolate sauce.
But you probably didn't know any of this because they don't think developers are their #1 customers. Someone there needs to watch Ballmer do his developer dance.
Crusoe's are cool. Transmeta was cool, too. Working there was like working down the hall from about a dozen John Carmacks. You could walk into any one of these offices and be blown away by what they were working on. They were crossing real-time translation and optimization bridges that Intel won't be getting to in years but will eventually have to face.
Microsoft learned long ago how important developers were. That should have been the main market to chase. Crusoe wasn't ready for the masses, not by a long shot. The performance is catching up with a vengeance with every new core, but they made so many promises and IPO'd on so much hype, that they entered the classic promise debt trap that so many dot-coms fell into, and their lofty marketing plan claiming that benchmarks are "wrong" (please!) and that it offers this brilliant power savings are just goofy.
Had they remained lean, not staffed up to 400+ people from the 150ish they had when I joined, and stayed quiet, humble, and in the service of developers until developers helped propel them to mass marketability, they would not be the laughing stock they are today.
Yes, they hoped to be faster than "native" x86 based computers by morphing to VLIW, but what they didn't realize was that there would be a terrible price in instruction bandwidth. They ended up with a lemon, made lemonade, then added red food coloring and called it wine.
If they exposed an API to CMS, I think they would be truly impressed by the tricks that independent developers could come up with to compress their own instruction stream to make the compression ratio competetive with x86 code footprint.
Do you realize that's really the main performance problem with Crusoe? The instruction bandwidth! On average, x86 instructions, because they're variable width and byte granularity, are 6X smaller than the average Crusoe instruction, which is made up of two or four 32-bit atoms in the current architecture.
OK, that's too darn bad, but it's the youngest surviving newcomer to the x86 market, and this is a solvable problem, and "with many eyeballs, all problems become shallow," once said a bright chap who ought to put his foot down and say it again.
Tablets: I made the demos that ran on tablets for their shows and IPO roadshow. They're cute, fun, no market for them yet, but again, something to get in the hands of developers so that they can make killer apps to create a market for tablets.